Somewhere on the vvv is a discussion about the cancelling effect of valvesprings : do they actually HELP turning the camshaft (when the valve is closing) with the force of the spring (meanwhile "helping" to turn the engine) or is it all waisted effort ?
How does the complete drivetrain of the camshafts act in this "help", having some backlash on some parts doesn't help much.
The singles Desmo had weaker springs than the spring-valves but still needed much effort to turn them manually.
I've seen heads where they were dropped and a Vtwin Imola cam was fitted (the "easy" way) without the smaller desmosprings.
Now they weld the head, reshape everything from chamber, ports to valve-angle and fit springs with a specially designed cam of course.
So : someone better grinds a BevelDesmocamshaft with the same timing/lift as the 860 Bevelspringvalvecam, and (everything else being equal) puts it on a dyno to get a reading .
I've let the springs out of my Imolacam Bevel racer, because it was difficult to fit them (didn't have that strange tool yet) without messing around with the shims. But setting the closer rocker at 0.00 mm is difficult and shimconsuming work.
On my streetbikes (where valves didn't need to be checked so often) they are in and working, eliminating the need to shim to the last few 0.0X's of a mm.
Some of the rockers had a fault in the curve, and binded just a bit when opening the valve.
Run them that way and rockers AND cam are trashed beyond repair.
So I ALWAYS turned them by hand during and after shimming to be sure they turned freely and if needed gave the valves a bit more play.
Carlo is right about springs helping eliminate the (very light) bouncing of the valve due to the clearance of the closing rocker, not so sure about they really getting rid of it.
When they ask me why I wrote this :
It were the little springs inside my head that made me do it

ciao
ducadini